U.S. patent number 11,280,860 [Application Number 17/168,274] was granted by the patent office on 2022-03-22 for method for generating mri rf pulse and device for the same.
This patent grant is currently assigned to SEOUL NATIONAL UNIVERSITY R&DB FOUNDATION. The grantee listed for this patent is SEOUL NATIONAL UNIVERSITY R&DB FOUNDATION. Invention is credited to Jongho Lee, Dongmyung Shin.
United States Patent |
11,280,860 |
Lee , et al. |
March 22, 2022 |
Method for generating MRI RF pulse and device for the same
Abstract
Disclosed is an MRI control signal providing method including
obtaining an initial control variable array including time-series
values of a control variable for controlling a spatial profile of
an induced magnetic field induced by an MRI scanner, obtaining
information about a desired spatial profile of the induced magnetic
field in the MRI scanner, calculating a differentiation array
obtainable by partially differentiating a predetermined function
with respect to the control variable, and calculating a scaled
array obtained by scaling the differentiation array with a
predetermined scaling factor, and generating an updated control
variable array from the initial control variable array by
subtracting values of the scaled array from values of the initial
control variable array.
Inventors: |
Lee; Jongho (Seoul,
KR), Shin; Dongmyung (Seoul, KR) |
Applicant: |
Name |
City |
State |
Country |
Type |
SEOUL NATIONAL UNIVERSITY R&DB FOUNDATION |
Seoul |
N/A |
KR |
|
|
Assignee: |
SEOUL NATIONAL UNIVERSITY R&DB
FOUNDATION (Seoul, KR)
|
Family
ID: |
80269487 |
Appl.
No.: |
17/168,274 |
Filed: |
February 5, 2021 |
Prior Publication Data
|
|
|
|
Document
Identifier |
Publication Date |
|
US 20220057461 A1 |
Feb 24, 2022 |
|
Foreign Application Priority Data
|
|
|
|
|
Aug 24, 2020 [KR] |
|
|
10-2020-0106569 |
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G01R
33/543 (20130101); G01R 33/3607 (20130101) |
Current International
Class: |
G01R
33/36 (20060101); G01R 33/54 (20060101) |
References Cited
[Referenced By]
U.S. Patent Documents
Primary Examiner: Hyder; G.M. A
Attorney, Agent or Firm: Mayer & Williams, PC Mayer;
Stuart H.
Claims
What is claimed is:
1. A method for driving an MRI device comprising an MRI scanner and
an MRI scanner control device, the method comprising: obtaining, by
the MRI scanner control device, a control variable array including
time-series values of a control variable for controlling a spatial
profile of an induced magnetic field induced by the MRI scanner;
calculating, by the MRI scanner control device, a differentiation
array obtainable by partially differentiating a predetermined
function with respect to the control variable, and calculating a
scaled array obtained by scaling the differentiation array with a
predetermined scaling factor; generating, by the MRI scanner
control device, an updated control variable array from the control
variable array by subtracting values of the scaled array from
values of the control variable array; and applying, by the MRI
scanner, a driving current generated using the time-series values
of the control variable included in the updated control variable
array to a coil included in the MRI scanner, wherein the
predetermined function receives, as input variables, a fixed
magnetic field provided by the MRI scanner, the control variable
array, and information about a desired spatial profile of the
induced magnetic field in the MRI scanner, wherein an output value
of the predetermined function is a distance between simulated
spatial profile values of the induced magnetic field and desired
spatial profile values of the induced magnetic field.
2. An MRI control signal providing method comprising: obtaining, by
a computing device, a control variable array including time-series
values of a control variable for controlling a spatial profile of
an induced magnetic field induced by an MRI scanner; calculating,
by the computing device, a differentiation array obtainable by
partially differentiating a predetermined function with respect to
the control variable, and calculating a scaled array obtained by
scaling the differentiation array with a predetermined scaling
factor; and generating, by the computing device, an updated control
variable array from the control variable array by subtracting
values of the scaled array from values of the control variable
array, wherein the predetermined function receives, as input
variables, a fixed magnetic field provided by the MRI scanner, the
control variable array, and information about a desired spatial
profile of the induced magnetic field in the MRI scanner, wherein
an output value of the predetermined function is a distance between
simulated spatial profile values of the induced magnetic field and
desired spatial profile values of the induced magnetic field.
3. The MRI control signal providing method of claim 2, wherein the
control variable comprises: a variable indicating an amplitude of
an RF pulse output from the MRI scanner; a variable indicating a
phase of the RF pulse; a variable indicating a value of a real part
of the RF pulse; a variable indicating a value of an imaginary part
of the RF pulse; a variable indicating a change rate of a magnitude
or intensity of a gradient magnetic field per unit distance along
an x-axis direction in a scan space of the MRI scanner; a variable
indicating the change rate of the magnitude or intensity of the
gradient magnetic field per unit distance along a y-axis direction
in the scan space; or a variable indicating the change rate of the
magnitude or intensity of the gradient magnetic field per unit
distance along a z-axis direction in the scan space.
4. The MRI control signal providing method of claim 2, wherein the
simulated spatial profile values of the induced magnetic field are
values of an induced magnetic field calculated by a simulation at a
set of voxels selected from among a plurality of voxels defined in
a scan space of the MRI scanner, and the desired spatial profile
values of the induced magnetic field are values of an induced
magnetic field predefined for the set of voxels selected.
5. The MRI control signal providing method of claim 4, wherein the
set of voxels are voxels selected according to a predetermined rule
from among all of the voxels defined in the scan space, and the
distance is a distance between component values indicating
components of a particular direction among the values of an induced
magnetic field calculated by the simulation and component values
indicating components of the particular direction among the values
of the induced magnetic field predefined.
6. The MRI control signal providing method of claim 2, wherein the
simulated spatial profile values of the induced magnetic field and
the desired spatial profile values of the induced magnetic field
comprise the same number of elements, and wherein a procedure of
calculating the distance comprises: calculating difference values
between the simulated spatial profile values of the induced
magnetic field and the desired spatial profile values of the
induced magnetic field corresponding thereto; calculating a square
of each of the difference values; and setting a value obtained by
adding up all of the calculated squares as the distance.
7. The MRI control signal providing method of claim 2, wherein the
method comprises a control signal updating process including the
obtaining, the calculating, and the generating, wherein the control
signal updating process is repeatedly executed until the distance
reaches a predetermined threshold value or less, and, when the
distance reaches the predetermined threshold value or less,
information about the control signal is provided to a control
device of the MRI scanner or a storage device readable by the
control device, and wherein each time the updating process is
executed, the updated control variable array replaces the control
variable array.
8. The MRI control signal providing method of claim 2, wherein the
calculating of the differentiation array is performed using an
automatic differentiation part including a computation graph with
the predetermined function as a target function.
9. The MRI control signal providing method of claim 2, further
comprising providing, by the computing device, information about
the updated control variable array to a control device of the MRI
scanner or a storage device readable by the control device as
information for generating a driving current of a coil of the MRI
scanner.
10. A computing device for controlling an MRI scanner, the
computing device comprising a communication interface and a
processing part, wherein the processing part is configured to:
obtain a plurality of control variable arrays each including
time-series values of a plurality of control variables for
controlling a spatial profile of an induced magnetic field induced
by the MRI scanner; calculate a plurality of differentiation arrays
obtainable by partially differentiating a predetermined function
with respect to each of the control variables, and calculate a
plurality of scaled arrays obtained by scaling each of the
differentiation arrays with a predetermined scaling factor; and
generate a plurality of updated control variable arrays from the
plurality of control variable arrays by subtracting values of each
of the scaled arrays from values of the corresponding control
variable array, wherein the predetermined function receives, as
input variables, a fixed magnetic field provided by the MRI
scanner, the plurality of control variable arrays, and information
about a desired spatial profile of the induced magnetic field in
the MRI scanner, wherein an output value of the predetermined
function is a distance between simulated spatial profile values of
the induced magnetic field and desired spatial profile values of
the induced magnetic field in the MRI scanner.
11. The computing device of claim 10, wherein the processing part
is configured to provide information about the plurality of updated
control variable arrays to a control device of the MRI scanner or a
storage device readable by the control device via the communication
interface as information for generating a driving current of coils
of the MRI scanner.
12. The computing device of claim 10, wherein the plurality of
control variables comprise: a variable indicating an amplitude of
an RF pulse output from the MRI scanner and a variable indicating a
phase of the RF pulse; or a variable indicating a value of a real
part of the RF pulse and a variable indicating an imaginary part of
the RF pulse.
13. The computing device of claim 10, wherein the plurality of
control variables comprise: a variable indicating a change rate of
a magnitude or intensity of a gradient magnetic field per unit
distance along an x-axis direction in a scan space of the MRI
scanner; a variable indicating the change rate of the magnitude or
intensity of the gradient magnetic field per unit distance along a
y-axis direction in the scan space; and a variable indicating the
change rate of the magnitude or intensity of the gradient magnetic
field per unit distance along a z-axis direction in the scan space.
Description
CROSS-REFERENCE TO RELATED APPLICATION
This application claims priority to Korean Patent Application No.
10-2020-0106569 filed on Aug. 24, 2020, and all the benefits
accruing therefrom under 35 U.S.C. .sctn. 119, the contents of
which are incorporated by reference in their entirety.
BACKGROUND
The present disclosure relates to a signal processing technique for
generating a signal used in an MRI scanner. The present disclosure
particularly relates to a technique for determining waveforms of an
RF pulse and gradient magnetic field over time.
An MRI scanner may provide, within a scan space in the MRI scanner,
a magnetic field determined by a vector sum of a fixed magnetic
field B0 that does not change according to a position and time, a
gradient magnetic field G.sub.x,y,z(t) that may change according to
time and may linearly vary in magnitude in a particular direction
in the space, and a variable magnetic field B.sub.RF provided since
an RF pulse is applied.
Due to the magnetic field provided by the MRI scanner, an induced
magnetic field may be formed in each position of a scan target
object disposed in the scan space. The induced magnetic field may
have different values according to each position in the scan space.
In the present disclosure, for example, a pattern of the induced
magnetic field according to x-axis, y-axis, and z-axis directions
may be referred to as a spatial profile. An operator of such an MRI
scanner may desire that the spatial profile of the induced magnetic
field have a desired shape.
SUMMARY
The present disclosure provides a technology of designing the
gradient magnetic field and RF pulse so as to achieve a preferred
spatial profile of an induced magnetic field, which is desired by
an operator of an MRI scanner.
FIGS. 1A-1F are diagrams for describing a pattern of a magnetic
field in a scan space of an MRI scanner to which an RF pulse is
applied.
As illustrated in the part (a) of FIG. 1, a fixed magnetic field B0
oriented in a first direction, for example, z-axis direction, may
be provided in the scan space of the MRI scanner. The fixed
magnetic field B0 constantly has the same magnitude at arbitrary
coordinates (x,y,z) in the scan space. Furthermore, the fixed
magnetic field B0 is constantly oriented in the first direction at
the arbitrary coordinates (x,y,z) in the scan space.
Furthermore, a gradient magnetic field G.sub.x,y,z(t) may be
additionally provided in the scan space of the MRI scanner. The
gradient magnetic field G.sub.x,y,z(t) may be constantly oriented
in the first direction at arbitrary spatial coordinates (x,y,z) in
the scan space. However, the gradient magnetic field G.sub.x,y,z(t)
may have a magnitude controlled according to a predefined rule at
each of spatial coordinates (x,y,z). The gradient magnetic field
G.sub.x,y,z(t) may have different magnitudes according to the
spatial coordinates (x,y,z).
Change patterns of the magnitude of the gradient magnetic field
G.sub.x,y,z(t) according to a space may be classified and defined
as three patterns.
That is, as illustrated in each of the part (b) of FIG. 1, the part
(c) of FIG. 1, and the part (d) of FIG. 1, the gradient magnetic
field may be controlled to linearly increase or decrease in
magnitude along x-axis, y-axis, and z-axis directions from an
origin point defined on the x-axis, an origin point defined on the
y-axis, and an origin point defined on the z-axis. Here, a change
rate of the magnitude or intensity of the gradient magnetic field
per unit distance along the x-axis direction may be referred to by
.DELTA.Gx, the change rate of the magnitude or intensity of the
gradient magnetic field per unit distance along the y-axis
direction may be referred to by .DELTA.Gy, and the change rate of
the magnitude or intensity of the gradient magnetic field per unit
distance along the z-axis direction may be referred to by
.DELTA.Gz.
Each of .DELTA.Gx, .DELTA.Gy, and .DELTA.Gz may change according to
time. Here, .DELTA.Gx, .DELTA.Gy, and .DELTA.Gz may be respectively
expressed as .DELTA.Gx(t), .DELTA.Gy(t), and .DELTA.Gz(t). Change
patterns of .DELTA.Gx(t), .DELTA.Gy(t), and .DELTA.Gz(t) according
to time have values settable in the MRI scanner.
It would be easily understood that when the change patterns of
.DELTA.Gx(t), .DELTA.Gy(t), and .DELTA.Gz(t) according to time are
defined, the magnitude of the gradient magnetic field
G.sub.x,y,z(t) at arbitrary spatial coordinates (x,y,z) in the scan
space and at an arbitrary time may be recognized.
An RF pulse having an amplitude and phase which change according to
time may be applied in the scan space. The part (e) of FIG. 1
exemplarily illustrates a change in the amplitude of the RF pulse
according to time, and the part (f) of FIG. 1 exemplarily
illustrates a change in the phase of the RF pulse according to
time. When the RF pulse is applied to the scan space, a variable
magnetic field B.sub.RF provided by the RF pulse may be
additionally provided to each of spatial coordinates (x,y,z) of the
scan space.
FIG. 2 is a diagram for describing a voxel magnetic field induced
at different spatial coordinates (x,y,z) in the scan space.
FIG. 3 illustrates voxels defined by dividing the scan space of the
MRI scanner into multiple sub spaces according to a predefined
rule, and examples of a desired spatial profile of an induced
magnetic field that is induced from voxels arranged in a particular
direction and a spatial profile of an induced magnetic field that
is actually induced when a schemed RF pulse and schemed gradient
magnetic field are applied. One cubic of a basic unit that may be
recognized from FIG. 3 represents one voxel.
Descriptions will be given with reference to both FIGS. 2 and
3.
As illustrated in the part (a) of FIG. 3, the scan space of the MRI
scanner may be divided into multiple sub spaces according to the
predefined rule, wherein each of the sub spaces may be referred to
as a voxel. That is, the scan space may be defined as one including
a plurality of voxels. A boundary shape of each voxel may be
arbitrary determined, but, in an embodiment, each voxel may be
defined in a cubic form as illustrated in the part (a) of FIG.
3.
As illustrated in the part (a) of FIG. 3, the scan space may be
defined as a set of voxels arranged adjacent to each other in a
three-dimensional matrix form. In this case, each voxel may be
expressed as V(p,q,r) or Vox(p,q,r) in order to distinguish each
voxel, where p, q, and r respectively denote an index given along
the x direction of the voxel, an index given along the y direction,
and an index given along the z direction.
When an MRI scan target object occupies only a portion of the scan
space, each voxel may include or may not include a portion of the
MRI scan target object. An example of the MRI scan target object
may include an organism or organic matter.
Since a single voxel may have a volume that is not zero, voxel
magnetic fields B.sub.Vox provided as a vector sum of the fixed
magnetic field B0, the gradient magnetic field G.sub.x,y,z(t), and
the variable magnetic field B.sub.RF provided by an RF pulse may
have different values at various positions in the single voxel.
However, when the single voxel is defined to have a sufficiently
small size, the values of the voxel magnetic fields B.sub.Vox
provided at each position in the single voxel may approximate to
the same value. Therefore, a voxel magnetic field provided by the
MRI scanner in a particular voxel may be expressed as a single
value B.sub.Vox. It is obvious that voxel magnetic fields provided
to different voxels may have different values.
A magnetic field provided to a single particular voxel by the MRI
scanner may be presented as a vector expressed as the intensity and
direction of the magnetic field. At the moment when a gradient
magnetic field and an RF pulse are provided in the scan space of
the MRI scanner, the voxel magnetic field B.sub.Vox provided to a
single particular voxel may be defined as a vector sum of the fixed
magnetic field B0, the gradient magnetic field G.sub.Vox provided
to the particular voxel, and the variable magnetic field
B.sub.RF,Vox provided to the particular voxel by the provided RF
pulse.
For example, as illustrated in the parts (a) and (b) of FIG. 2, a
first gradient magnetic field G.sub.Vox1 and a second gradient
magnetic field G.sub.Vox2 provided to a first voxel Vox1 and a
second voxel Vox2 may be the same or different, and a first
variable magnetic field B.sub.RFVox1 and a second variable magnetic
field B.sub.RFVox2 provided by the RF pulse may also be the same or
different. Therefore, as illustrated in the parts (a) and (b) of
FIG. 2, a first voxel magnetic field B.sub.Vox1 and a second voxel
magnetic field B.sub.Vox2 may be different.
When the MRI scan target object is present in a single voxel, and
when the MRI scan target object includes water molecules, the
induced magnetic field M.sub.Vox may be formed on the MRI scan
target object due to the voxel magnetic field B.sub.Vox provided to
the single voxel. That is, the induced magnetic field M.sub.Vox
that is induced and formed in a particular voxel may be determined
by the voxel magnetic field B.sub.Vox provided to the particular
voxel by the MRI scanner.
This induced magnetic field M.sub.Vox may be defined for each
voxel, and may be formed to have the same value or different values
for each voxel.
An induced magnetic field formed for a given single voxel may be
expressed as a vector. For example, the induced magnetic field
M.sub.Vox may be decomposed into an x component Mx.sub.Vox, a y
component My.sub.Vox, and a z component Mz.sub.Vox so as to be
presented. The x component Mx.sub.Vox may represent an x-axis
direction component of the induced magnetic field, the y component
My.sub.Vox may represent a y-axis direction component of the
induced magnetic field, and the z component Mz.sub.Vox may
represent a z-axis direction component.
Here, in the part (a) of FIG. 3, for example, in a state in which a
voxel index q of the y axis and a voxel index r of the z axis are
fixed to particular values, particular components of the induced
magnetic fields formed on a row of voxels arranged along the x
axis, for example, the z component Mz.sub.Vox, may be
considered.
Here, as illustrated in the parts (b) and (c) of FIG. 3, when
considering a graph space in which a horizontal axis is defined as
the x axis (or p axis), and a vertical axis represents a magnitude
of the z component Mz.sub.Vox, a single graph may be obtained by
displaying, in the graph space, the z components Mz.sub.Vox
obtained for the row of voxels arranged along the axis
direction.
This graph may be referred to as a spatial profile of an induced
magnetic field that is formed on an MRI scan target object by a
given MRI scanner.
Although the parts (b) and (c) of FIG. 3 illustrate that the graphs
are continuous along the horizontal axes, the horizontal axes,
which are for distinguishing a limited number of voxels, may be
actually discrete. Thus, the graphs illustrated in the parts (b)
and (c) of FIG. 3 may be construed as being presented by
interpolating points illustrated for each voxel.
The spatial profile of the induced magnetic field that is formed on
the MRI scan target object due to the applied RF pulse may be
presented in various aspects.
In an example, spatial distribution of characteristic values of the
induced magnetic field obtained for an arbitrary group of voxels
selected according to one aspect may be defined as a spatial
profile of the induced magnetic field. For more specific example,
spatial distribution of component values of a particular direction,
for example, the z-axis direction, of the induced magnetic field
obtained for a group of voxels selected according to one aspect may
be defined as a spatial profile of the induced magnetic field.
The arbitrary group of voxels selected according to one aspect, for
example, may be configured with a plurality of voxels arbitrarily
selected from the plurality of voxels illustrated in FIG. 3 without
particularly limiting values of the x axis, y axis and z axis.
Alternatively, in a state in which values of two axes among the x
axis, y axis and z axis are fixed, the arbitrary group of voxels
selected according to one aspect may be configured with a plurality
of voxels selected from the other one axis.
In another example, spatial distribution of component values of a
particular direction of the induced magnetic field obtained for a
row of selected voxels arranged in a particular axis direction may
be defined as a spatial profile of the induced magnetic field.
Here, the particular axis direction may be the x, y, or z
direction.
The parts (b) and (c) of FIG. 3 illustrate spatial profiles
presented along the x direction.
The parts (b) and (c) of FIG. 3 exemplarily illustrate spatial
profiles of induced magnetic fields for a group of voxels
continuously selected along a straight line of the x axis (p axis).
However, unlike this illustration, such a spatial profile may also
be defined by a set of arbitrary voxels spaced apart from each
other in the scan space. However, in this case, it is difficult to
display the spatial profile on paper.
Furthermore, a component of a particular direction of the induced
magnetic field M.sub.Vox may be an x-axis direction component
Mx.sub.Vox of the induced magnetic field M.sub.Vox, a y-axis
direction component My.sub.Vox of the induced magnetic field
M.sub.Vox, or a z-axis direction component Mz.sub.Vox of the
induced magnetic field M.sub.Vox. The parts (b) and (c) of FIG. 3
exemplarily illustrates the z-axis direction component Mz.sub.Vox
of the induced magnetic field M.sub.Vox.
Here, the spatial profile of the induced magnetic field M.sub.Vox
is determined by the voxel magnetic field B.sub.Vox applied to each
voxel, wherein the voxel magnetic field B.sub.Vox is determined by
a gradient magnetic field and RF pulse. Therefore, it would be
understood that the spatial profile of the induced magnetic field
M.sub.Vox varies according to specific shapes of the gradient
magnetic field and RF pulse.
The part (b) of FIG. 3 illustrates an example of a desired spatial
profile of an induced magnetic field presented to satisfy an
application desired by an operator of an MRI scanner, and the part
(c) of FIG. 3 illustrates an example of an undesired spatial
profile of an induced magnetic field simulated when a given RF
pulse is applied.
According one aspect of the present invention, an MRI control
signal providing method may be provided. The method comprises,
obtaining, by a computing device, a control variable array
including time-series values of a control variable for controlling
a spatial profile of an induced magnetic field induced by an MRI
scanner; calculating, by the computing device, a differentiation
array obtainable by partially differentiating a predetermined
function with respect to the control variable, and calculating a
scaled array obtained by scaling the differentiation array with a
predetermined scaling factor; and generating, by the computing
device, an updated control variable array from the control variable
array by subtracting values of the scaled array from values of the
control variable array, wherein the predetermined function
receives, as input variables, a fixed magnetic field provided by
the MRI scanner, the control variable array, and information about
a desired spatial profile of the induced magnetic field in the MRI
scanner, wherein an output value of the predetermined function is a
distance between simulated spatial profile values of the induced
magnetic field and desired spatial profile values of the induced
magnetic field.
Here, the control variable may comprises a variable indicating an
amplitude of an RF pulse output from the MRI scanner; a variable
indicating a phase of the RF pulse; a variable indicating a value
of a real part of the RF pulse; a variable indicating a value of an
imaginary part of the RF pulse; a variable indicating a change rate
of a magnitude or intensity of a gradient magnetic field per unit
distance along an x-axis direction in a scan space of the MRI
scanner; a variable indicating the change rate of the magnitude or
intensity of the gradient magnetic field per unit distance along a
y-axis direction in the scan space; or a variable indicating the
change rate of the magnitude or intensity of the gradient magnetic
field per unit distance along a z-axis direction in the scan
space.
In an embodiment of the present invention, a variable indicating
the amplitude of an RF pulse output from the MRI scanner and a
variable indicating the phase of the RF pulse may be used as
control variables. However, in another embodiment, instead of such
variables, a variable indicating a value of a real part of the RF
pulse and a variable indicating an imaginary part of the RF pulse
may be used.
Here, the plurality of control variables may comprise a variable
indicating an amplitude of an RF pulse output from the MRI scanner
and a variable indicating a phase of the RF pulse; or a variable
indicating a value of a real part of the RF pulse and a variable
indicating an imaginary part of the RF pulse.
Here, the simulated spatial profile values of the induced magnetic
field may be values of an induced magnetic field calculated by a
simulation at a set of voxels selected from among a plurality of
voxels defined in a scan space of the MRI scanner, and the desired
spatial profile values of the induced magnetic field may be values
of an induced magnetic field predefined for the set of voxels
selected.
In an embodiment of the present invention, a set of voxels selected
from among a plurality of voxels defined in the scan space of the
MRI scanner may be used. Here, for example, the selected set of
voxels may be a row of voxels arranged continuously along the
x-axis, a row of voxels arranged continuously along the y-axis, or
a row of voxels arranged continuously along the z-axis.
Alternatively, the selected set of voxels may be a set of voxels
having a particular x value and present on a y-z plane, a set of
voxels having a particular y value and present on an x-z plane, or
a set of voxels having a particular z value and present on an x-y
plane.
Alternatively, the selected set of voxels may be a set of voxels
arbitrarily selected in an x-y-z space or selected according to a
predetermined rule. Here, the selected set of voxels may not be
present on a single line and may not be present on a single
plane.
Alternatively, the set of voxels may be a single row of voxels
arranged along an axis direction selected from among x-axis,
y-axis, and z-axis directions defined in the scan space.
Here, the set of voxels may be voxels selected according to a
predetermined rule from among all of the voxels defined in the scan
space, and the distance may be a distance between component values
indicating components of a particular direction among the values of
an induced magnetic field calculated by the simulation and
component values indicating components of the particular direction
among the values of the induced magnetic field predefined.
Here, the simulated spatial profile values of the induced magnetic
field and the desired spatial profile values of the induced
magnetic field may comprise the same number of elements, and
wherein a procedure of calculating the distance may comprises:
calculating difference values between the simulated spatial profile
values of the induced magnetic field and the desired spatial
profile values of the induced magnetic field corresponding thereto;
calculating a square of each of the difference values; and setting
a value obtained by adding up all of the calculated squares as the
distance.
Here, the method may comprise a control signal updating process
including the obtaining, the calculating, and the generating,
wherein the control signal updating process is repeatedly executed
until the distance reaches a predetermined threshold value or less,
and, when the distance reaches the predetermined threshold value or
less, information about the control signal is provided to a control
device of the MRI scanner or a storage device readable by the
control device, and wherein each time the updating process is
executed, the updated control variable array replaces the control
variable array.
Here, the calculating of the differentiation array may be performed
using an automatic differentiation part including a computation
graph with the predetermined function as a target function.
An MRI control signal providing method provided according to
another aspect of the present invention includes: obtaining, by a
computing device, a plurality of control variable arrays each
including time-series values of a plurality of control variables
for controlling a spatial profile of an induced magnetic field
induced by the MRI scanner; calculating, by the computing device, a
plurality of differentiation arrays obtainable by partially
differentiating a predetermined function with respect to each of
the control variables, and calculating a plurality of scaled arrays
obtained by scaling each of the differentiation arrays with a
predetermined scaling factor; and generating, by the computing
device, a plurality of updated control variable arrays from the
plurality of control variable arrays by subtracting values of each
of the scaled arrays from values of the corresponding control
variable array. Here, the predetermined function receives, as input
variables, a fixed magnetic field provided by the MRI scanner, the
plurality of control variable arrays, and information about a
desired spatial profile of the induced magnetic field in the MRI
scanner, and an output value of the predetermined function is a
distance between simulated spatial profile values of the induced
magnetic field and desired spatial profile values of the induced
magnetic field.
The method may further include providing, by the computing device,
information about the plurality of updated control variable arrays
to a control device of the MRI scanner or a storage device readable
by the control device as information for generating a driving
current of coils of the MRI scanner.
Here, the plurality of control variables may include a variable
indicating an amplitude of an RF pulse output from the MRI scanner
and a variable indicating a phase of the RF pulse, or a variable
indicating a value of a real part of the RF pulse and a variable
indicating an imaginary part of the RF pulse.
Here, the plurality of control variables may include a variable
indicating a change rate of a magnitude or intensity of a gradient
magnetic field per unit distance along an x-axis direction in a
scan space of the MRI scanner, a variable indicating the change
rate of the magnitude or intensity of the gradient magnetic field
per unit distance along a y-axis direction in the scan space, and a
variable indicating the change rate of the magnitude or intensity
of the gradient magnetic field per unit distance along a z-axis
direction in the scan space.
According to one aspect of the present invention, a computing
device for controlling an MRI scanner can be provided. The
computing device comprises a communication interface and a
processing part. The processing part is configured to: obtain a
plurality of control variable arrays each including time-series
values of a plurality of control variables for controlling a
spatial profile of an induced magnetic field induced by the MRI
scanner; calculate a plurality of differentiation arrays obtainable
by partially differentiating a predetermined function with respect
to each of the control variables, and calculate a plurality of
scaled arrays obtained by scaling each of the differentiation
arrays with a predetermined scaling factor; and generate a
plurality of updated control variable arrays from the plurality of
control variable arrays by subtracting values of each of the scaled
arrays from values of the corresponding control variable array,
wherein the predetermined function receives, as input variables, a
fixed magnetic field provided by the MRI scanner, the plurality of
control variable arrays, and information about a desired spatial
profile of the induced magnetic field in the MRI scanner, wherein
an output value of the predetermined function is a distance between
simulated spatial profile values of the induced magnetic field and
desired spatial profile values of the induced magnetic field in the
MRI scanner.
Here, the processing part may be configured to provide information
about the plurality of updated control variable arrays to a control
device of the MRI scanner or a storage device readable by the
control device via the communication interface as information for
generating a driving current of coils of the MRI scanner.
According to another aspect of the present invention, a computing
device including a communication interface and a processing part
may be provided. Here, the processing part is configured to obtain
a plurality of control variable arrays each including time-series
values of a plurality of control variables for controlling a
spatial profile of an induced magnetic field induced by an MRI
scanner. Furthermore, the processing part is configured to
calculate a plurality of differentiation arrays obtainable by
partially differentiating a predetermined function with respect to
each of the control variables, and calculate a plurality of scaled
arrays obtained by scaling each of the differentiation arrays with
a predetermined scaling factor. Furthermore, the processing part is
configured to generate a plurality of updated control variable
arrays from the plurality of control variable arrays by subtracting
values of each of the scaled arrays from values of the
corresponding control variable array. Here, the predetermined
function receives, as input variables, a fixed magnetic field
provided by the MRI scanner, the plurality of control variable
arrays, and information about a desired spatial profile of the
induced magnetic field in the MRI scanner. Furthermore, an output
value of the predetermined function is a distance between simulated
spatial profile values of the induced magnetic field and desired
spatial profile values of the induced magnetic field in the MRI
scanner.
The computing device may be one for controlling an MRI scanner.
According to another aspect of the present invention, a method for
driving an MRI device including an MRI scanner and an MRI scanner
control device may be provided. The method may include obtaining,
by the MRI scanner control device, a control variable array
including time-series values of a control variable for controlling
a spatial profile of an induced magnetic field induced by the MRI
scanner. Furthermore, the method may include calculating, by the
MRI scanner control device, a differentiation array obtainable by
partially differentiating a predetermined function with respect to
the control variable, and calculating a scaled array obtained by
scaling the differentiation array with a predetermined scaling
factor. Furthermore, the method may include generating, by the MRI
scanner control device, an updated control variable array from the
control variable array by subtracting values of the scaled array
from values of the control variable array. Furthermore, the method
may include applying, by the MRI scanner, a driving current
generated using the time-series values of the control variable
included in the updated control variable array to a coil included
in the MRI scanner.
Here, the predetermined function may receive, as input variables, a
fixed magnetic field provided by the MRI scanner, the control
variable array, and information about a desired spatial profile of
the induced magnetic field in the MRI scanner. Furthermore, an
output value of the predetermined function may be a distance
between simulated spatial profile values of the induced magnetic
field and desired spatial profile values of the induced magnetic
field.
BRIEF DESCRIPTION OF THE DRAWINGS
Exemplary embodiments can be understood in more detail from the
following description taken in conjunction with the accompanying
drawings, in which:
FIG. 1A, FIG. 1B, FIG. 1C, FIG. 1D, FIG. 1E and FIG. 1F are
diagrams for describing a magnetic field in a scan space of an MRI
scanner to which an RF pulse is applied;
FIG. 2A and FIG. 2B are diagrams for describing a voxel magnetic
field induced at different points in a scan space;
FIG. 3A, FIG. 3B and FIG. 3C illustrate voxels defined by dividing
a scan space of an MRI scanner into multiple sub spaces according
to a predefined rule, and examples of a desired profile of an
induced magnetic field that is induced from voxels arranged in a
particular direction and a profile of an induced magnetic field
that is actually induced when a schemed RF pulse and schemed
gradient magnetic field are applied;
FIG. 4 is a block diagram illustrating a method of generating an RF
pulse and a gradient magnetic field, provided according to one
embodiment of the present invention;
FIG. 5 is a block diagram illustrating a method of generating an RF
pulse and a gradient magnetic field, provided according to another
embodiment of the present invention;
FIG. 6 is an exemplary diagram for describing a structure of an
automatic differentiation part;
FIG. 7 illustrates input data and a target function of an automatic
differentiation part used in the present invention;
FIG. 8 is a diagram illustrating a control signal updating method
provided according to an embodiment of the present invention;
FIG. 9 is a diagram illustrating a control signal updating method
provided according to another embodiment of the present
invention;
FIG. 10A and FIG. 10B show a flowchart illustrating a control
signal providing method provided according to an embodiment of the
present invention;
FIG. 11A and FIG. 11B show a flowchart illustrating a control
signal providing method provided according to an embodiment of the
present invention;
FIG. 12 illustrates a relationship between a computing device, an
MRI scanner, and an MRI scanner control device provided according
to an embodiment of the present invention;
FIG. 13 is a diagram illustrating a configuration of an MRI device
provided according to an embodiment of the present invention;
FIG. 14 illustrates a method for driving an MRI device provided
according to an embodiment of the present invention; and
FIG. 15 illustrates a method for driving an MRI device provided
according to another embodiment of the present invention.
DETAILED DESCRIPTION OF EMBODIMENTS
Embodiments of the present invention will be described with
reference to the accompanying drawings. However, the present
invention is not limited to the embodiments described herein, and
may be implemented in various different forms. The terminology used
herein is not for limiting the scope of the present invention but
for describing the embodiments. Furthermore, the singular forms
used herein include the plural forms as well, unless otherwise
indicated.
Hereinafter, examples of an RF pulse and two methods of expressing
an RF pulse, addressed in the present invention, will be described
with reference to Equation 1 and Equation 2.
An RF pulse may be a signal having a complex value that changes
according to time. As illustrated in the part (e) of FIG. 1, the RF
pulse may have an amplitude that changes according to time. As
illustrated in the part (f) of FIG. 1, the RF pulse may have a
phase that changes according to time.
As shown in Equation 1 and Equation 2 below, the RF pulse may be
expressed as RF[n] indicating values sampled on a time axis at a
predetermined period, where n may be a natural number.
.function..function..times..function..times..times..function..times..time-
s..times..function..times..times. ##EQU00001##
.function..function..times..times..times..function..times..times..functio-
n..times..times..times..times..times..function..times..times.
##EQU00002##
When n has at least a particular value in Equation 1 and Equation
2, var1[n] and var2[n] may be constantly 0. That is, RF[n] may be a
pulse having a finite length.
The RF pulse RF[n] may represent a signal emitted from an RF pulse
radiating coil provided to an MRI scanner.
Alternatively, the RF pulse RF[n] may represent a signal of an RF
pulse driving current applied to the coil or a control signal for
controlling a flow of the driving current.
The RF pulse driving current may be provided as continuous values,
but it would be understood that the control signal for controlling
the RF pulse driving current may be provided as discrete values.
Therefore, as shown in Equation 1 and Equation 2, the symbol n
which indicates discrete time may be used instead of the symbol t
which indicates continuous time.
In an embodiment represented by Equation 1, each sampled RF pulse
RF[n] may include a real part Re[n] and an imaginary part Im[n].
Here, the real part Re[n] may be regarded as a first control
variable var1[n], and the imaginary part Im[n] may be regarded as a
second control variable var2[n]. An array of RF pulses RF[n]
sampled at different time points may include an array of the first
control variables var1[n] and an array of the second control
variables var2[n].
In another embodiment represented by Equation 2, each sampled RF
pulse RF[n] may include an amplitude part Amp[n] and a phase part
phase[n]. Here, the amplitude part Amp[n] may be regarded as the
first control variable var1[n], and the phase part phase[n] may be
regarded as the second control variable var2[n]. An array of RF
pulses RF[n] sampled at different time points may include an array
of the first control variables var1[n] and an array of the second
control variables var2[n].
With regard to above-described Equation 1 and Equation 2, it may be
recognized that the array of RF pulses RF[n] sampled at different
time points may include an array of the first control variables
var1[n] and an array of the second control variables var2[n]
regardless of in which form the RF pulse RF[n] is expressed.
Hereinafter, the RF pulse RF[n] will be described as following
Equation 2 in a preferred embodiment of the present invention, but
the concept of the present invention described below may also be
applied even when the RF pulse RF[n] is presented in the form of
Equation 1.
A profile of the amplitude of the RF pulse according to time may be
determined by an array including variables var1[n] (=Amp[n])
defined for different values of n.
Furthermore, a profile of the phase of the RF pulse according to
time may be determined by an array including variables var2[n]
(=phase[n]) defined for different values of n.
A gradient magnetic field G.sub.x,y,z(t) may be provided by a group
of coils arranged in the MRI scanner. Here, .DELTA.Gx(t),
.DELTA.Gy(t), and .DELTA.Gz(t) may be expressed as discrete time
.DELTA.Gx[n], .DELTA.Gy[n], and .DELTA.Gz[n].
As shown in Equation 3, Equation 4, and Equation 5, .DELTA.Gx[n],
.DELTA.Gy[n], and .DELTA.Gz[n] may be respectively expressed as
var3[n], var4[n], and var5[n]. .DELTA.Gx[n]=var3[n] [Equation 3]
.DELTA.Gy[n]=var4[n] [Equation 4] .DELTA.Gz[n]=var5[n] [Equation
5]
Here, a profile of a change rate of a magnitude or intensity of the
gradient magnetic field per unit distance along the x-axis
direction according to time may be determined by an array including
variables var3[n] (=.DELTA.Gx[n]) defined for different values of
n. Here, n is a parameter indicating time.
Furthermore, the profile of the change rate of the magnitude or
intensity of the gradient magnetic field per unit distance along
the y-axis direction according to time may be determined by an
array including variables var4[n] (=.DELTA.Gy[n]) defined for
different values of n.
Furthermore, the profile of the change rate of the magnitude or
intensity of the gradient magnetic field per unit distance along
the z-axis direction according to time may be determined by an
array including variables var5[n] (=.DELTA.Gz[n]) defined for
different values of n.
FIG. 4 is a block diagram illustrating a method of generating an RF
pulse and a gradient magnetic field, provided according to one
embodiment of the present invention.
Each of the blocks illustrated in FIG. 4 may be executed on a
computing device. The computing device may be provided with a
simulator 120 implemented as software. Alternatively, the simulator
120 may be provided as a dedicated hardware device. The simulator
120 may receive six types of variables from a DB 110. The DB 110
may be provided inside or outside the computing device.
The six types of variables are described as below.
A first variable V1 is a first array Array{var1.iter.sub.k}
including the first control variables var1[n] that constitute the
above-described RF pulse.
A second variable V2 is a second array Array{var2.iter.sub.k}
including the second control variables var2[n] that constitute the
above-described RF pulse.
A third variable V3 is a third array Array{var3.iter.sub.k}
including variables var3[n] (=.DELTA.Gx[n]) indicating the profile
of the change rate of the magnitude or intensity of the gradient
magnetic field per unit distance along the x-axis direction
according to time.
A fourth variable V4 is a fourth array Array{var4.iter.sub.k}
including variables var4[n] (=.DELTA.Gy[n]) indicating the profile
of the change rate of the magnitude or intensity of the gradient
magnetic field per unit distance along the y-axis direction
according to time.
A fifth variable V5 is a fifth array Array{var5.iter.sub.k}
including variables var5[n] (=.DELTA.Gz[n]) indicating the profile
of the change rate of the magnitude or intensity of the gradient
magnetic field per unit distance along the z-axis direction
according to time.
A sixth variable V6 is a basic magnetic field B0 provided by the
MRI scanner in which the RF pulse is provided.
The simulator 120 may receive the six types of variables to
simulate and calculate an induced magnetic field M.sub.p,q,r that
is induced from each of voxels defined in the scan space of the MRI
scanner in which the RF pulse RF[n], the basic magnetic field B0,
and the gradient magnetic field G.sub.p,q,r(n) are provided. A
specific method for this simulation may be selected from among
conventional techniques. The present invention is not limited by
the specific method for the simulation.
Each simulated induced magnetic field M.sub.simulated(p,q,r)
induced on each voxel may be defined as a vector value. Therefore,
each simulated induced magnetic field M.sub.simulated(p,q,r) may
include an x component Mx.sub.simulated(p,q,r), a y component
My.sub.simulated(p,q,r), and z component
Mz.sub.simulated(p,q,r).
Since a plurality of voxels are provided along each of the x axis,
y axis and z axis, the simulated induced magnetic field
M.sub.simulated(p,q,r) may be calculated for each voxel so as to be
provided in plurality. Therefore, the x components of the simulated
induced magnetic field M.sub.simulated(p,q,r) may be provided for
each voxel, and this may be referred to as a first S-induced
magnetic field array Array{Mx.sub.simulated(p,q,r)}. Furthermore,
the y components of the simulated induced magnetic field
M.sub.simulated(p,q,r) may be provided for each voxel, and this may
be referred to as a second S-induced magnetic field array
Array{My.sub.simulated(p,q,r)}. Furthermore, the z components of
the simulated induced magnetic field M.sub.simulated(p,q,r) may be
provided for each voxel, and this may be referred to as a third
S-induced magnetic field array Array(Mz.sub.simulated(p,q,r)).
The first S-induced magnetic field array
Array{Mx.sub.simulated(p,q,r)}, the second S-induced magnetic field
array Array{My.sub.simulated(p,q,r)}, and the third S-induced
magnetic field array Array{Mz.sub.simulated(p,q,r)} may be
expressed in a vector form and simply referred to as simulated
induced magnetic field array Array{M.sub.simulated(p,q,r)}.
The simulator 120 may provide a portion or entirety of the
simulated induced magnetic field array
Array{M.sub.simulated(p,q,r)} to a distance calculator 130.
In the embodiment illustrated in FIG. 4, it is assumed that the
simulated induced magnetic field array
Array{M.sub.simulated(p,q,r)} provided to the distance calculator
130 is the third S-induced magnetic field array
Array{Mz.sub.simulated(p,q,r)}.
However, in another embodiment, the simulated induced magnetic
field array provided to the distance calculator 130 may be the
first S-induced magnetic field array Array{Mx.sub.simulated(p,q,r)}
or the second S-induced magnetic field array
Array{My.sub.simulated(p,q,r)}.
Alternatively, in another embodiment, an induced magnetic field
array provided to the distance calculator 130 may be the simulated
induced magnetic field array Array{M.sub.simulated(p,q,r)}.
The computing device may receive, from the DB 110, a spatial
profile of a desired induced magnetic field, i.e., a desired
induced magnetic field array Array{M.sub.DESIRED(p,q,r)}.
The desired induced magnetic field array
Array{M.sub.DESIRED(p,q,r)} may be stored in the DB 110 as a
seventh variable.
The seventh variable V7 may include the desired induced magnetic
field array Array{M.sub.DESIRED(p,q,r)} including magnitudes of
desired induced magnetic fields desired to be induced from each of
a plurality of voxels. The desired induced magnetic field array
Array{M.sub.DESIRED(p,q,r)} may include, for each of x, y, z
components thereof, a first D-induced magnetic field array
Array{Mx.sub.DESIRED(p,q,r)}, a second D-induced magnetic field
array Array{My.sub.DESIRED(p,q,r)}, and a third D-induced magnetic
field array Array{M.sub.DESIRED(p,q,r)}.
In the embodiment illustrated in FIG. 4, the third D-induced
magnetic field array Array{Mz.sub.DESIRED(p,q,r)} is exemplarily
provided to the distance calculator 130.
However, in another embodiment, a desired induced magnetic field
array provided to the distance calculator 130 may be the first
D-induced magnetic field array Array{Mx.sub.DESIRED(p,q,r)} or the
second D-induced magnetic field array
Array{My.sub.DESIRED(p,q,r)}.
In another embodiment, an induced magnetic field array provided to
the distance calculator 130 may be the desired induced magnetic
field array Array{M.sub.DESIRED(p,q,r)}.
The distance calculator 130 may calculate a distance between the
simulated induced magnetic field array
Array{M.sub.simulated(p,q,r)} provided from the simulator 120 and
the desired induced magnetic field array
Array{M.sub.DESIRED(p,q,r)} provided from the DB 110.
Here, a set of voxels corresponding to the simulated induced
magnetic field array Array{M.sub.simulated(p,q,r)} may be the same
as a set of voxels corresponding to the desired induced magnetic
field array Array{M.sub.DESIRED(p,q,r)}. For example, the simulated
induced magnetic field array Array{M.sub.simulated(p,q,r)} and the
desired induced magnetic field array Array{M.sub.DESIRED(p,q,r)}
may be an array including a plurality of voxels selected along a p
axis in a state in which q and r among indices p, q, and r for
distinguishing voxels are fixed to particular values.
FIG. 4 illustrates an example in which the distance calculator 130
calculates a distance between the third S-induced magnetic field
array Array{Mz.sub.simulated(p,q,r)} which is a simulated induced
magnetic field array and the third D-induced magnetic field array
Array{Mz.sub.DESIRED(p,q,r)} which is a desired induced magnetic
field array.
The third S-induced magnetic field array
Array{Mz.sub.simulated(p,q,r)} and the third D-induced magnetic
field array Array{Mz.sub.DESIRED(p,q,r)} may have the same
size.
Here, the above-described distance may be defined as a value
obtained by calculating differences between corresponding elements
of the third S-induced magnetic field array
Array{Mz.sub.simulated(p,q,r)} and the third D-induced magnetic
field array Array{Mz.sub.DESIRED(p,q,r)} and then adding up squares
of the differences.
Therefore, during a process of calculating the above-described
distance, all of the first to seventh variables V1, V2, V3, V4, V5,
V6, and V7 are used. Thus, in an embodiment of the present
invention, the above-described distance may be expressed as
Equation 6 below. Distance=F(V1,V2,V3,V4,V5,V6,V7) [Equation 6]
The above-described distance may be defined in various ways other
than the above-described method, and the present invention is not
limited by a specific method of defining the distance.
A differentiator 140 may calculate values by partially
differentiating the calculated distance with respect to a first
control variable var1, a second control variable var2, a third
control variable var3, a fourth control variable var4, and a fifth
control variable var5. That is, the differentiator 140 may
calculate values by partially differentiating a function F (V1, V2,
V3, V4, V5, V6, V7), which has the first to seventh variables V1,
V2, V3, V4, V5, V6, and V7 as input variables, with respect to the
first control variable var1, the second control variable var2, the
third control variable var3, the fourth control variable var4, and
the fifth control variable var5.
As a result, the differentiator 140 may output a first
differentiation array
Array{(.differential.F/.differential.var1).iter.sub.k}, a second
differentiation array
Array{(.differential.F/.differential.var2).iter.sub.k}, a third
differentiation array
Array{(.differential.F/.differential.var3).iter.sub.k}, a fourth
differentiation array
Array{(.differential.F/.differential.var4).iter.sub.k}, and a fifth
differentiation array
Array{(.differential.F/.differential.var5).iter.sub.k} including
the values obtained by partially differentiating the function with
respect to the first control variable var1, the second control
variable var2, the third control variable var3, the fourth control
variable var4, and the fifth control variable var5.
An RF pulse profile updating part 150 may update and generate an RF
pulse using the first array Array{var1.iter.sub.k}, the second
array Array{var2.iter.sub.k}, the first differentiation array
Array{(.differential.F/.differential.var1).iter.sub.k}, and the
second differentiation array
Array{(.differential.F/.differential.var2).iter.sub.k}. To this
end, following Equation 7 may be used.
var1[n].iter.sub.k+1=var1[n].iter.sub.k+.alpha..quadrature..differential.-
F/.differential.var1[n].iter.sub.k
var2[n].iter.sub.k+1=var2[n].iter.sub.k+.alpha..quadrature..differential.-
F/.differential.var2[n].iter.sub.k [Equation 7]
That is, a value obtained by scaling the first differentiation
array Array{(.differential.F/.differential.var1).iter.sub.k} with a
predetermined scaling factor .alpha. may be added to the first
array Array{var1.iter.sub.k} constituting the RF pulse.sub.k
RF.sub.k[n], and a resultant value may be used as a first array
Array{var1.iter.sub.k+1} constituting a new RF pulse.sub.k+1
RF.sub.k+1[n]. Furthermore, a value obtained by scaling the second
differentiation array
Array{(.differential.F/.differential.var2).iter.sub.k} with the
predetermined scaling factor .alpha. may be added to the second
array Array{var2.iter.sub.k} constituting the RF pulse.sub.k
RF.sub.k[n], and a resultant value may be used as a second array
Array{var2.iter.sub.k+1} constituting the new RF pulse.sub.k+1
RF.sub.k+1[n].
Through this process, a relationship between a value of the newly
updated RF pulse.sub.k+1 RF.sub.k+1[n] sampled at a particular time
point and a value of the RF pulse.sub.k RF.sub.k[n] sampled at the
particular time point is expressed Equation 8.
.function..times..times..function..times..times..times..quadrature..times-
..times..times..function..times..times..function..alpha..times..quadrature-
..times..differential..differential..times..times..function..times..times.-
.times..quadrature..times..times..times..function..alpha..times..quadratur-
e..times..differential..differential..times..times..function..function..ti-
mes..times..times..quadrature..times..function..times..times.
##EQU00003##
When the above-described process is repeated for the new RF
pulse.sub.k+1 RF.sub.k+1[n], a new RF pulse.sub.k+2 RF.sub.k+2[n]
may be obtained.
A gradient magnetic field profile updating part 160 may update and
generate a gradient magnetic field profile using the third array
Array{var3.iter.sub.k}, the fourth array Array{var4.iter.sub.k},
the fifth array Array{var5.iter.sub.k}, the third differentiation
array Array{(.differential.F/.differential.var3).iter.sub.k}, the
fourth differentiation array
Array{(.differential.F/.differential.var4).iter.sub.k}, and the
fifth differentiation array
Array{(.differential.F/.differential.var5).iter.sub.k}. To this
end, following Equation 9 may be used.
var3[n].iter.sub.k+1=var3[n].iter.sub.k+.beta..quadrature..differential.F-
/.differential.var3[n].iter.sub.k
var4[n].iter.sub.k+1=var4[n].iter.sub.k+.beta..quadrature..differential.F-
/.differential.var3[n].iter.sub.k
var5[n].iter.sub.k+1=var5[n].iter.sub.k+.beta..quadrature..differential.F-
/.differential.var3[n].iter.sub.k [Equation 9]
That is, a new third array Array{var3.iter.sub.k+1} may be
generated by adding a value obtained by scaling the third
differentiation array
Array{(.differential.F/.differential.var3).iter.sub.k} with a
predetermined scaling factor .beta. to the third array
Array{var3.iter.sub.k} including variables var3[n] (=.DELTA.Gx[n])
that indicate the profile of the change rate of the magnitude or
intensity of the gradient magnetic field per unit distance along
the x-axis direction according to time.
Furthermore, a new fourth array Array{var4.iter.sub.k+1} may be
generated by adding a value obtained by scaling the fourth
differentiation array
Array{(.differential.F/.differential.var4).iter.sub.k} with the
predetermined scaling factor .beta. to the fourth array
Array{var4.iter.sub.k} including variables var4[n] (=.DELTA.Gy[n])
that indicate the profile of the change rate of the magnitude or
intensity of the gradient magnetic field per unit distance along
the y-axis direction according to time.
Furthermore, a new fifth array Array{var5.iter.sub.k+1} may be
generated by adding a value obtained by scaling the fifth
differentiation array
Array{(.differential.F/.differential.var5).iter.sub.k} with the
predetermined scaling factor .beta. to the fifth array
Array{var5.iter.sub.k} including variables var5[n] (=.DELTA.Gz[n])
that indicate the profile of the change rate of the magnitude or
intensity of the gradient magnetic field per unit distance along
the z-axis direction according to time.
The scaling factor .beta. may have a value equal to or different
from the scaling factor .alpha..
As described above, a gradient magnetic field profile.sub.k (V3,
V4, V5) expressed by reference sign k may be updated by the
gradient magnetic field profile updating part 160 into a new
gradient magnetic field profile.sub.k+1 (V3', V4', V5') expressed
by reference sign k+1.
When the above-described process is repeated for the new gradient
magnetic field profile.sub.k+1, a new gradient magnetic field
profile.sub.k+2 may be obtained.
This process may be continuously repeated, and, when the
above-described distance decreases to a preset value or less, the
RF pulse updating process and the gradient magnetic field profile
updating process may be ended.
A final RF pulse and a final gradient magnetic field profile
obtained when the RF pulse updating process and the gradient
magnetic field profile updating process are ended may be stored in
the computing device.
The final RF pulse and the final gradient magnetic field profile
may be stored in a volatile storage or non-volatile storage of the
computing device.
In an embodiment, information about the stored final RF pulse and
final gradient magnetic field profile may be transmitted to the MRI
scanner via a local communication network or a network
communication network. Here, the local communication network may
represent a LAN cable, a local communication cable, or the like for
connecting the computing device and the MRI scanner to each other.
Furthermore, the network communication network may be a
communication network including a service server of an Internet
service provider (ISP) for connecting the computing device to a
remotely located MRI scanner.
In another embodiment, information about the stored final RF pulse
and final gradient magnetic field profile may be transmitted, via
the local communication network or the network communication
network, to a second computing device which operates the MRI
scanner. The second computing device may be a server. The second
computing device may drive, based on a command input thereto, a
current driver for generating an RF pulse and a gradient magnetic
field of the MRI scanner using the information about the final RF
pulse and final gradient magnetic field profile. The second
computing device may also be a control computing device integrated
with the MRI scanner.
FIG. 5 is a block diagram illustrating a method of generating an RF
pulse and a gradient magnetic field, provided according to another
embodiment of the present invention.
FIG. 5 illustrates a modification example of the embodiment of FIG.
4, and, in this example, functions of the simulator 120, the
distance calculator 130, and the differentiator 140 of FIG. 4 are
replaced with an automatic differentiation part 170.
FIG. 6 is an exemplary diagram for describing a structure of an
automatic differentiation part.
FIG. 7 illustrates input data and a target function of an automatic
differentiation part used in the present invention.
The automatic differentiation part 170 which is a known technology
may include a computation graph 59 therein. The computation graph
59 may include a graph of a tree structure including a plurality of
nodes and links, and is configured to implement a predetermined
target function which uses variables input to the computation graph
59 as input variables.
For example, when the input variables are x.sub.1 and x.sub.2, and
the target function f(x.sub.1,
x.sub.2)=ln(x.sub.1)+x.sub.1x.sub.2-sin(x.sub.2) is given, the
target function may be implemented by configuring the computation
graph 59 illustrated in FIG. 6.
Therefore, as illustrated in FIG. 7, the first to seventh variables
V1, V2, V3, V4, V5, V6, and V7 may be used as the input variables
of the computation graph 59 illustrated in FIG. 5, and the function
F(V1, V2, V3, V4, V5, V6, V7) indicating the above-described
distance may be used as the target function.
Here, a specific method for designing an internal structure of the
computation graph 59 to implement the target function from the
input variables depends on a designer's decision. The present
invention is not limited by the specific method for designing the
internal structure of the computation graph 59.
The automatic differentiation part 170 is configured to calculate
partially differentiated values of each of the variables for the
target function by executing a forward mode or a reverse mode.
Therefore, it may be understood that the automatic differentiation
part 170 of FIG. 5 may substitute for the functions of the
simulator 120, the distance calculator 130, and the differentiator
140 of FIG. 4.
FIG. 8 is a diagram illustrating a control signal updating method
provided according to an embodiment of the present invention.
FIG. 8 illustrates a modification example of the embodiment of FIG.
4, and, in this example, only one variable array Vx among five
variable arrays V1, V2, V3, V4, and V5 is used.
FIG. 9 is a diagram illustrating a control signal updating method
provided according to another embodiment of the present
invention.
FIG. 9 illustrates a modification example of the embodiment of FIG.
5, and, in this example, only one variable array Vx among five
variable arrays V1, V2, V3, V4, and V5 is used.
FIG. 4 and FIG. 5 illustrate examples of updating all of the five
variable arrays V1, V2, V3, V4, and V5, but, even if only one of
the arrays is updated, the spatial profile of a simulated induced
magnetic field may be approximated to the spatial profile of a
desired induced magnetic field. Therefore, the aspect of the
present invention includes the concept of updating any one array or
two or more selected arrays among the five variable arrays V1, V2,
V3, V4, and V5.
The five variable arrays V1, V2, V3, V4, and V5 share a common role
of a control signal for determining the spatial profile of an
induced magnetic field. Thus, each of the five variable arrays V1,
V2, V3, V4, and V5 may be referred to as a control variable array
Vx or control signal Vx including time-series values of a control
variable for determining the spatial profile of an induced magnetic
field.
It would be easily understood that the control signal updating part
180 illustrated in FIGS. 8 and 9 collectively refers to the RF
pulse profile updating part 150 and the gradient magnetic field
profile updating part 160 illustrated in FIGS. 4 and 5.
FIG. 10A and FIG. 10B show a flowchart illustrating a control
signal providing method provided according to an embodiment of the
present invention.
Referring to FIG. 10A and FIG. 10B, the control signal providing
method may include the following operations.
In operation S10, a computing device may obtain, from an external
or internal storage device of the computing device accessible by
the computing device, a control variable array Vx including
time-series values of a control variable for controlling the
spatial profile of an induced magnetic field induced by an MRI
scanner, and may obtain information about a desired spatial profile
of the induced magnetic field in the MRI scanner.
Here, the control variable, for example, may be a variable
indicating the amplitude of an RF pulse, a variable indicating the
phase of the RF pulse, a variable indicating the change rate of the
magnitude or intensity of a gradient magnetic field per unit
distance along the x-axis direction, a variable indicating the
change rate of the magnitude or intensity of the gradient magnetic
field per unit distance along the y-axis direction, or a variable
indicating the change rate of the magnitude or intensity of the
gradient magnetic field per unit distance along the z-axis
direction.
In operation S20, the computing device may calculate a
differentiation array that may be obtained by partially
differentiating a predetermined function with respect to the
control variable.
Here, the predetermined function may receive, as input variables,
(1) the fixed magnetic field B0 provided by the MRI scanner, (2)
the control variable array Vx, and (3) the information about the
desired spatial profile of the induced magnetic field in the MRI
scanner.
Here, spatial profiles of the fixed magnetic field B0, a gradient
magnetic field generated based on the control variable array Vx,
and an induced magnetic field induced by a variable magnetic field
caused by an RF pulse generated based on the control variable array
Vx may be obtained through simulation.
Furthermore, an output value of the predetermined function may be a
distance between simulated spatial profile values of the induced
magnetic field and desired spatial profile values of the induced
magnetic field.
In operation S30, the computing device may generate an updated
control variable array Vx' from the control variable array Vx by
subtracting values of an array obtained by scaling the
differentiation array with a predetermined scaling factor from
values of the control variable array.
In operation S40, the computing device may provide information
about the updated control variable array Vx' to a storage device
readable by the MRI scanner or a control device of the MRI scanner
as information for generating a driving current of an RF coil of
the MRI scanner and/or a driving current of a coil which generates
a gradient magnetic field.
Alternatively, in operation S41 instead of operation S40, the
computing device may repeat operation S10, operation S20, and
operation S30 by replacing the updated control variable array Vx'
with the control variable array Vx of operation S10.
FIG. 11A and FIG. 11B show a flowchart illustrating a control
signal providing method provided according to an embodiment of the
present invention.
Referring to FIG. 11A and FIG. 11B, the control signal providing
method may include the following operations.
In operation S110, a computing device may obtain, from an internal
or external storage device of the computing device accessible by
the computing device, an initial control variable array Vx0
including time-series values of a control variable var_x for
controlling the spatial profile of an induced magnetic field
induced by an MRI scanner, and may obtain information V7 about a
desired spatial profile of the induced magnetic field in the MRI
scanner.
Here, the control variable var_x, for example, may be a variable
var1 indicating the amplitude of an RF pulse, a variable var2
indicating the phase of the RF pulse, a variable var3 indicating
the change rate of the magnitude or intensity of a gradient
magnetic field per unit distance along the x-axis direction, a
variable var4 indicating the change rate of the magnitude or
intensity of the gradient magnetic field per unit distance along
the y-axis direction, or a variable var5 indicating the change rate
of the magnitude or intensity of the gradient magnetic field per
unit distance along the z-axis direction.
The initial control variable array Vx0 may include all or selected
portion of the above-described various types of control
variables.
In operation S120, the computing device may replace a value of a
predetermined input array Vx with a value of the initial control
variable array Vx0.
The input array Vx may be one used to use an output value of the
given predetermined function described below.
Here, the given function may receive, as input variables, (1) the
fixed magnetic field B0 provided by the MRI scanner, (2) the input
array Vx, and (3) the information V7 about the desired spatial
profile. Furthermore, the output value of the function may be a
distance between simulated spatial profile values of the induced
magnetic field and the desired spatial profile values V7.
In operation S200, the computing device may calculate the output
value of the function by inputting the input array Vx to the
function.
In operation S210, the computing device may determine whether the
output value of the function is equal to or less than a
predetermined threshold value, and the process may proceed to
operation S300 if the output value is equal to or less than the
threshold value, or may proceed to operation S220 if the output
value is greater than the threshold value.
In operation S220, the computing device may calculate a
differentiation array that may be obtained by partially
differentiating the function with respect to the control
variable.
In operation S230, the computing device may generate an updated
control variable array Vx' by subtracting or adding values of a
scaled array obtained by scaling the differentiation array with a
predetermined scaling factor from or to values of the input array
Vx.
In operation S240, the computing device may replace a value of the
input array Vx with a value of the updated control variable array
Vx'. Thereafter, the process may return to operation S200.
In operation S300, the computing device may provide information
about the input array Vx to a storage device readable by the MRI
scanner or a control device of the MRI scanner as information for
generating a driving current of an RF coil of the MRI scanner
and/or a driving current of a coil which generates a gradient
magnetic field.
According to the method of the embodiment illustrated in FIG. 11A
and FIG. 11B, the input array Vx which starts from the initial
control variable array Vx0 is continuously updated until the output
value of the predetermined function, i.e., the distance value,
converges to a certain level or lower.
The flowchart illustrated in FIG. 10A and FIG. 10B may be
considered to show only a last repetition procedure of the updating
procedure illustrated in FIG. 11A and FIG. 11B.
FIG. 12 illustrates a relationship between a computing device, an
MRI scanner, and an MRI scanner control device provided according
to an embodiment of the present invention.
A computing device 1 may include a processing part 12, a storage
part 15, a database 110, and a communication interface 13. In an
embodiment, the database 110 may be provided separately from the
computing device 1.
The communication interface 13 may include a network connecting
device 132 and a media controller 131. The media controller 131 may
be, for example, a USB controller.
The computing device 1 may be connected to an MRI scanner control
device 3 via LAN/WAN, short-range communication network, or
baseband connection. The MRI scanner control device 3 may be a
device for controlling operation of an MRI scanner 2.
In an embodiment, the computing device 1 may be substantially
integrated with the MRI scanner control device 3.
Program codes to be executed by the processing part 12 may be
recorded in the storage part 15.
The simulator 120, the distance calculator 130, the differentiator
140, the RF pulse profile updating part 150, the gradient magnetic
field profile updating part 160, the automatic differentiation part
170, and the control signal updating part 180 may be functional
software modules executed by the program codes in the processing
part 12.
The information about the control variable array Vx' updated and
generated in the processing part 12 may be provided to the MRI
scanner control device 3 via the network connecting device 132, or
may be stored in a predetermined storage device via the media
controller 131.
FIG. 13 is a diagram illustrating a configuration of an MRI device
provided according to an embodiment of the present invention.
FIG. 14 illustrates a method for driving an MRI device provided
according to an embodiment of the present invention.
Descriptions will be given with reference to both FIGS. 13 and
14.
An MRI device 100 may include the MRI scanner 2 and the MRI scanner
control device 3.
The MRI device 100 may include an RF coil 210 for radiating an RF
pulse and a gradient magnetic field generation coil 220 used to
generate a gradient magnetic field. The RF coil 210 and the
gradient magnetic field generation coil 220 may be collectively
referred to as a coil.
The MRI scanner control device 3 may include the processing part
12, the communication interface 13, the database 110, the simulator
120, the distance calculator 130, the differentiator 140, the RF
pulse profile updating part 150, the gradient magnetic field
profile updating part 160, the automatic differentiation part 170,
and the control signal updating part 180.
The processing part 12 may control operation of the communication
interface 13, the database 110, the simulator 120, the distance
calculator 130, the differentiator 140, the RF pulse profile
updating part 150, the gradient magnetic field profile updating
part 160, the automatic differentiation part 170, and the control
signal updating part 180 in compliance with the above-described
embodiments of the present invention.
The communication interface 13 may perform a function of
transmitting information 190 about a predetermined control signal
generated by the MRI scanner control device 3 to the MRI scanner 2
and receiving an MRI signal 240 obtained by the MRI scanner 2 from
a scan target object 300.
A method for driving the MRI device 100 including the MRI scanner 2
and the MRI scanner control device 3 according to an embodiment of
the present invention may include operations S410, S420, and S430
described below.
In operation S410, a prescribed control signal updating process
provided according to an embodiment of the present invention may be
repeatedly performed until a prescribed condition is satisfied.
The control signal updating process may include operations S411,
S412, and S413 described below.
In operation S411, the MRI scanner control device 3 may obtain a
control variable array including time-series values of a control
variable for controlling the spatial profile of an induced magnetic
field induced by the MRI scanner 2.
In operation S412, the MRI scanner control device 3 may calculate a
differentiation array that may be obtained by partially
differentiating a predetermined function with respect to the
control variable, and may calculate a scaled array by scaling the
differentiation array with a predetermined scaling factor.
Here, the predetermined function may receive, as input variables, a
fixed magnetic field provided by the MRI scanner 2, the control
variable array, and information about a desired spatial profile of
the induced magnetic field in the MRI scanner 2.
Furthermore, an output value of the predetermined function may be a
distance between simulated spatial profile values of the induced
magnetic field and desired spatial profile values of the induced
magnetic field.
In operation S413, the MRI scanner control device 3 may generate an
updated control variable array from the control variable array by
subtracting values of the scaled array from values of the control
variable array.
Each time the updating process is performed, the updated control
variable array may replace the control variable array.
Here, the prescribed condition of operation S410 may be a condition
that the distance between simulated spatial profile values of the
induced magnetic field and desired spatial profile values of the
induced magnetic field reach a prescribed threshold value or
less
Thereafter, in operation S420, the MRI scanner control device 3 may
provide, when the prescribed condition is satisfied and the
updating process is completed, the information 190 about the
control signal to the MRI scanner 2.
In operation S430, the MRI scanner 2 may apply, to a coil included
in the MRI scanner, a driving current generated based on the
information 190 about the control signal. That is, the MRI scanner
2 may apply, to the coil included in the MRI scanner, a driving
current generated using the time-series values of the control
variable included in the finally updated control variable array.
Here, the coil may be the RF coil 210 and/or the gradient magnetic
field generation coil 220.
In operation S440, the MRI scanner 2 may transfer, to the MRI
scanner control device 3, the MRI signal 240 received from the scan
target object 300 due to electromagnetic waves emitted from the
coil.
FIG. 15 illustrates a method for driving an MRI device provided
according to another embodiment of the present invention.
Descriptions will be given with reference to both FIGS. 13 and
15.
A method for driving the MRI device 100 including the MRI scanner 2
and the MRI scanner control device 3 according to another
embodiment of the present invention may include the following
operations.
In operation S510, the MRI scanner control device 3 may obtain a
control variable array including time-series values of a control
variable for controlling the spatial profile of an induced magnetic
field induced by the MRI scanner 2.
In operation S520, the MRI scanner control device 3 may calculate a
differentiation array that may be obtained by partially
differentiating a predetermined function with respect to the
control variable, and may calculate a scaled array by scaling the
differentiation array with a predetermined scaling factor.
Here, the predetermined function may receive, as input variables, a
fixed magnetic field provided by the MRI scanner 2, the control
variable array, and information about a desired spatial profile of
the induced magnetic field in the MRI scanner 2.
Furthermore, an output value of the predetermined function may be a
distance between simulated spatial profile values of the induced
magnetic field and desired spatial profile values of the induced
magnetic field.
In operation S530, the MRI scanner control device 3 may generate an
updated control variable array from the control variable array by
subtracting values of the scaled array from values of the control
variable array.
In operation S540, the MRI scanner 2 may apply, to the coils 210
and 220 included in the MRI scanner 2, a driving current generated
using the time-series values of the control variable included in
the updated control variable array.
In the present disclosure, p, q, and r may be parameters for
discretely expressing x, y, and z respectively.
The present invention may provide a technology of designing the
gradient magnetic field and RF pulse so as to achieve a desired
spatial profile of an induced magnetic field which is desired by a
user.
Although the present invention relates to a signal processing
technology, results of signal processing and signal conversion
according to the present invention may be used in controlling the
operation of an MRI scanner, thereby bringing about technical
improvement in the technical field of MRI scanning.
Those skilled in the art could easily make various alterations or
modifications to the above-mentioned embodiments of the present
invention without departing the essential characteristics of the
present invention. The claims that do not refer to each other may
be combined with each other within the scope of understanding of
the present disclosure.
* * * * *